270 likes | 504 Views
DMSO Workshop on Composable M&S 7-9-2002 Two Aspects of Composability: Lexicon and Theory Mikel D. Petty, Ph.D., Chief Scientist Virginia Modeling, Analysis and Simulation Center Old Dominion University mpetty@vmasc.odu.edu 757-686-6210. Presentation outline Composability lexicon
E N D
DMSO Workshop on Composable M&S 7-9-2002 Two Aspects of Composability: Lexicon and Theory Mikel D. Petty, Ph.D., Chief Scientist Virginia Modeling, Analysis and Simulation Center Old Dominion University mpetty@vmasc.odu.edu 757-686-6210
Presentation outline Composability lexicon Towards a theory of composability Conclusions and references
Same word, different meanings Different meanings for “composability” - Occurs in both documents and discussion - Noted previously [Page, 1998] - Found in review of composability literature [Petty, 2002] - Experienced at composability meeting The ability to rapidly configure, initialize, and test an exercise by logically assembling a simulation from a pool of reusable components [JSIMS, 1997]. The ability to create, configure, initialize, test, and validate an exercise by logically assembling a unique simulation execution from a pool of reusable system components in order to meet a specific set of objectives [Harkrider, 1999]. The ability to build new things from existing pieces [Pratt, 1999]. The ability to compose models/modules across a variety of application domains, levels of resolution and time scales [Kasputis, 2000].
Proposed common definition Composability is the capability to select and assemble simulation components in various combinations into simulation systems to satisfy specific user requirements. Component Repository Simulation B 1 Simulation A 3 1 2 8 3 3 22 52 . . . N N
Levels of composability Different meanings often differ in “level” - Level determined by component and composition - Different sets of levels identified in literature [JSIMS, 1997] [Page, 1998] [Petty, 2002] [Post, 2002] (explicitly) [U.S. Army, 1998] [Courtemanche, 2002] [Henderson, 2002] (implicitly) Components Composition Example Applications Events Unified Endeavor Federates Federations Joint Training Confederation Packages Simulations JSIMS Parameters Simulations JSIMS Modules Executables OneSAF Models Composite models OneSAF Data Databases SEDRIS Entities Military units WARSIM Behaviors Composite behaviors ModSAF
Application level - Applications composed into events - Applications: simulations, C4I systems, networks, communications systems, auxiliary software - aka “event” level [Post, 2002] Federate level - Federates composed into federations - Persistent federations, federates may change - aka “federation” level [Post, 2002] Package level - Packages composed into simulations [JSIMS, 1997] - Packages: sets of models forming consistent subset of battlespace [Page, 1998]
Parameter level - Parameters used to configure existing simulations - aka “simulation” level [JSIMS, 1997] Module level - Software modules composed into executables - Result may be federate or standalone simulation - Example: OneSAF [U.S. Army, 1998] [Courtemanche, 2002] Model level - Models composed into composite models - Example: sub-systems composed into entities - Example: OneSAF [Henderson, 2002] - aka“object”level[Post, 2002],“component”level[JSIMS, 1997]
Data level - Data sets composed into databases - Example: SEDRIS Entity level - Platforms/entities composed into military units - May be hierarchical, multi-leveled - aka “federate” level [Post, 2002] Behavior level - Behaviors composed into composite behaviors - Behaviors executed autonomously, e.g., CGF systems - Examples: FSMs [Calder, 1993], flow diagrams [Peters, 2002]
Composability and interoperability Interoperability is the ability of different simulations, connected in a distributed simulation, to meaningfully collaborate to simulate a common scenario or virtual world. Composability = interoperability? No. - Interoperability is necessary but not sufficient for composability - Composability requires interoperability - Interoperability possible without composability, i.e., without ability to combine and recombine, reuse - Example: Platform Proto-Federation [Harkrider, 1997] - Example: MV-22 OPEVAL Federation [Huntt, 2000]
Two types of composability Engineering composability - aka “syntactic” composability [Pratt, 1999] [Ceranowicz, 2002] - Ability to integrate composable components; APIs, parameter passing, data access, timing, ... Modeling composability - aka “semantic” composability [Pratt, 1999] [Ceranowicz, 2002] - Ability of composed models to meaningfully represent object; data content, assumptions, ... Analogous interoperability types - Engineering composability ~ technical interoperability - Modeling composability ~ substantive interoperability [Dahmann, 1999]
Introduction Much work on engineering composability - Dynamically loadable modules[Franceschini, 1999] - Autonomous behaviors[von der Lippe, 2002] [Peters, 2002] - OneSAF models[Henderson, 2002] Seek to address modeling composability - Develop and apply formal theory of composability - Some previous work [Overstreet, 1982] [Page, 1999] Project getting started - Dissertation, ODU M&S Ph.D. program - Eric W. Weisel - Topic still under development; current thinking here
Models Models are basis for the theory Official definition - Model: A physical, mathematical, or otherwise logical representation of a system, entity, phenomenon, or process [DOD, 1996] [DOD, 1998] - Assumption of validity? Formal definition for composability theory - Model: A computable function - Function: Relates each input to unique output - Computable: Calculable by finite procedure - No assumption of validity
h f g x g(f(x)) f(x) h(x) = g(f(x)) Linking model semantics, formal theory Validity links semantics and theory - Model semantics are established as validity - Models may or may not be valid - Is validity preserved when composing valid models? - That question may be answerable formally - Suppose models f(x) and g(x) are valid - Is composite model h(x) = g(f(x)) valid? - For what inputs x is composite model h(x) valid?
Domain Range xX f(x) Y f X Y y = f(x) Validity of composite models When is a composite model valid? Functions - Map domain (input) to range (output) - Exactly one output for any given input
Domain Range xX f(x) Y f Valid Valid xX f(x) Y X Y y = f(x) Validity and models as functions - Model valid for a subset of its domain - Model result valid in a subset of its range - Valid output produced from valid input
Range f Domain g Domain Range xX f(x) Y yY g(y) Z f g Valid Valid Valid Valid xX f(x) Y yY g(y) Z X Y Z z = g(f(x)) What is a composite model? - Models are functions - Composite models are composite functions When is a composite model valid? - When valid ranges and domains intersect
Validity How is validity determined initially? Option 1: meta-property of models - Determined outside formal system by validation - “Model f is valid for inputs X X” Option 2: formal property of models - Branch of mathematical logic: “model theory” - Relations between sentences in formal language, interpretations which make them true [Hodges, 2000] - Describe model and real-world axiomatically, establish validity inside formal system Then preserve validity in compositions
Simulation What is a simulation? Official definition - Simulation: A method for implementing a model over time [DOD, 1996] [DOD, 1998] Formal definition for composability theory - Simulation: A sequence of executions of a model i0 i1 i2 i3 Input values f f f f m1 m2 m0 m3 m4 Memory values Output values o1 o2 o3 o4 Steps Adapted from [Halbwachs, 1991] 1 2 3 4
Objectives Questions that may be resolved formally - Can the validity of a composition be checked algorithmically, given the validity of its components? - If yes, what is the algorithm, and what is its computational complexity? - What conditions must be met to preserve validity in a composition? - Canvaliditybeestablishedformallywithmodeltheory? - Are existing simulation formalisms (e.g., DEVS) equivalent to computable functions? - What is the computational complexity of selecting components to compose? [Page, 1999]
Conclusions Lexicon - Composability has several related meanings - Differences often depend on level, unit of composition - Common notions: interoperable, reusable, and recombinable components Theory - Modeling composability can be approached formally - Validity is link between semantics and theory - Computability theory and model theory are tools
References [Barrow, 1992] J. D. Barrow, Pi in the Sky: Counting, Thinking, and Being, Little, Brown, and Company, Boston MA, 1992. [Calder, 1993] R. B. Calder, J. E. Smith, A. J. Courtemanche, J. M. F. Mar, and A. Z. Ceranowicz, “ModSAF Behavior Simulation and Control”, Procedings of the Third Conference on Computer Generated Forces and Behavioral Representation, Orlando FL, March 17-19 1993, pp. 347-356. [Ceranowicz, 2002] A. Z. Ceranowicz, “Composability Wrapup”, Electronic mail, June 10 2002. [Courtemanche, 2002] A. J. Courtemanche and R. L. Wittman, “OneSAF: A Product Line Approach for a Next-Generation CGF”, Proceedings of the Eleventh Conference on Computer-Generated Forces and Behavior Representation, Orlando FL, May 7-9 2002, pp. 349-361. [Dahmann, 1999] J. S. Dahmann, “High Level Architecture Interoperability Challenges”, Presentation at the NATO Modeling & Simulation Conference, Norfolk VA, October 25-29 1999. [Davis, 1994] M. D. Davis, R. Sigal, and E. J. Weyuker, Computability, Complexity, and Languages: Fundamentals of Theoretical Computer Science, Second Edition, Morgan Kaufmann, San Diego, 1994. [DOD, 1996] DOD Instruction 5000.61, M&S VV&A, April 1996. [DOD, 1998] DOD Directive 5000.59, M&S Management, January 1998. [Franceschini, 1999] D. Franceschini, J. Zimmerman, and G. McCulley, “CGF System Composability through Dynamically Loadable Modules”, Proceedings of the Eighth Conference on Computer Generated Forces and Behavioral Representation, Orlando FL, May 11-13 1999, pp. 341-347. [Halbwachs, 1991] N. Halbwachs, P. Caspi, P. Raymond, and D. Pilaud, “The synchronous data flow programming language LUSTRE”, Proceedings of the IEEE, Vol. 79, Iss. 9, September 1991, pp. 1305-1320.
[Harkrider, 1997] S. M. Harkrider and M. D. Petty, “Results of the High Level Architecture Platform Proto- Federation Experiment”, Proceedings of the 8th International Training and Education Conference, Lausanne Switzerland, April 22-25 1997. [Harkrider, 1999] S. M. Harkrider and W. H. Lunceford, “Modeling and Simulation Composability” Proceedings of the 1999 Interservice/Industry Training, Simulation and Education Conference, Orlando FL, November 29 1999-December 2 1999. [Henderson, 2002] C. Henderson and A. Rodriguez, “Modeling in OneSAF”, Proceedings of the Eleventh Conference on Computer-Generated Forces and Behavior Representation, Orlando FL, May 7-9 2002, pp. 337-347. [Hodges, 2000] W. Hodges, “Model Theory”, in Concise Routledge Encyclopedia of Philosophy, Routledge, London, 2000. [Hu, 1969] S. Hu, Elementary Functions and Coordinate Geometry, Markham, Chicago, 1969. [Huntt, 2000] L. Huntt, A. Markowich, and L. Michelletti, “Modeling and Simulation Augments V-22 Operational Testing”, Proceedings of the 2000 Interservice/Industry Training, Simulation and Education Conference, November 27-30 2000, Orlando FL, pp. 945-953. [JSIMS, 1997] JSIMS Composability Task Force, “JSIMS Composability Task Force Final Report”, 1997. [Kasputis, 2000] S. Kasputis and H. C. Ng, “Composable simulations”, Proceedings of the 2000 Winter Simulation Conference, Orlando FL, December 10-13 2000, pp. 1577-1584. [Lloyd, 2000] E. A. Lloyd, “Models”, in in Concise Routledge Encyclopedia of Philosophy, Routledge, London, 2000. [Overstreet, 1982] C. M. Overstreet, Model Specification and Analysis for Discrete Event Simulation, Ph.D. Dissertation, Department of Computer Science, Virginia Polytechnic Institute and State University, Blacksburg VA, 1982.
[Page, 1998] E. H. Page and J. M. Opper, “Theory and Practice in User-Composable Simulation Systems”, Presentation for DARPA Advance Simulation Technology Thrust, October 30 1998. [Page, 1999] E. H. Page and J. M. Opper, “Observations on the Complexity of Composable Simulation”,Proceedings of the 1999 Winter Simulation Conference, Phoenix AZ, December 5-8 1999, pp. 553-560. [Peters, 2002] S. D. Peters, N. D. LaVine, L. Napravnik, and D. M. Lyons, “Composable Behaviors in an Entity Based Simulation”, Proceedings of the Spring 2002 Simulation Interoperability Workshop, Orlando FL, March 10-15 2002. [Petty, 2002] M. D. Petty, “A Composability Lexicon”, Unpublished manuscript, July 4 2002. [Post, 2002] G. M. Post, “J9 Composability Summary Comments”, Electronic mail, June 12 2002. [Pratt, 1999] D. R. Pratt, L. C. Ragusa, and S. von der Lippe, “Composability as an Architecture Driver”,Proceedings of the 1999 Interservice/Industry Training, Simulation and Education Conference, Orlando FL, November 29 1999-December 2 1999. [U. S. Army, 1998] United States Army, One Semi-Automated Forces Operational Requirements Document, Version 1.1, Online at http://www-leav.army.mil/nsc/stow/saf/onesaf/onesaf.htm/, August 21 1998. [von der Lippe, 2000] S. von der Lippe, J. S. McCormack, and M. Kalphat, “Embracing Temporal Relations and Command and Control in Composable Behavior Technologies”, Proceedings of the Ninth Conference on Computer Generated Forces and Behavioral Representation, Orlando FL, May 16-18 2000, pp. 183-192.